Patients with mental illness are two and a half times more likely to be victims of homicide than people in the general population, according to a national study examining the characteristics of homicide victims across England and Wales, published in The Lancet Psychiatry journal.

Homicides committed by patients with mental illness have received much media attention, but patients' risk of being victims of homicide and their relationship to the perpetrators has rarely been examined.

In this study, the National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (NCI) examined data on the victims and perpetrators of all homicides in England and Wales between January, 2003 and December, 2005.

The researchers found that during the 3-year study period, 1496 people were victims of homicide, and 6% (90) of them had been under the care of mental health services in the year before their death. A third (29) of these patient victims were killed by other patients with mental illness.

In 23 homicides in which the victim was a mental health patient killed by another mental health patient, the victim and the perpetrator were known to each other either as partners (9, 35%), family members (4, 15%), or acquaintances (10, 38%). In 21 of these 23 cases, both the victims and perpetrators were undergoing treatment at the same National Health Service Trust.

Alcohol and drug misuse (victims 66%, perpetrators 93%) and a history of violence (victims 24%, perpetrators 24%) were common among both patient victims and perpetrators. The study also found that in the 3 years to 2005, 213 mental health patients were convicted of homicide - accounting for 12% of all homicide convictions.

"Historically, society has been more concerned about the risk of patients committing violence than the vulnerability of patients to violent acts", explains study leader Professor Louis Appleby from The University of Manchester in the UK. "However, our findings show that specialist mental health providers in England and Wales can expect one of their patients to be the victim of homicide roughly every 2 years."*

He adds, "Assessing patients for risk of suicide and violence is common practice, but screening for risk of becoming victims of violence is not. Understanding that a patient's risk can depend on the environment they are in - for example their use of alcohol or drugs, or their contact with patients with a history of violence - and properly assessing these risk factors should become a key part of clinical care plans."*

Writing in a linked Comment, Alyssa Rheingold from the Medical University of South Carolina in the USA says, "Research suggests that risk factors among patient victims such as substance use, low socioeconomic status, type of psychopathology, and engagement in behaviours that increase risk could be targeted. Further exploration of these individual characteristics, their interactions, and their contribution to the risk of homicide is necessary. However, a full understanding of risk factors for homicide will be difficult to achieve in view of the potential effects of individual, situational-level, and community factors. Research must be undertaken to inform the development of a more comprehensive model to address the multifaceted manner in which individual and environmental characteristics contribute to risk."